The Blemishes of Modern Society? Acne Prevalence in the Dogon of Mali Christine E

325 original Evolution, Medicine, and Public Health [2016] pp. 325–337 research doi:10.1093/emph/eow027 article

The blemishes of modern society? Acne prevalence in the Dogon of Mali Christine E. Campbell1 and Beverly I. Strassmann*,1

1Department of Anthropology & Institute for Social Research, University of Michigan, Ann Arbor, MI, USA *Corresponding author. Department of Anthropology & Research Center for Group Dynamics, University of Michigan, 426 Thompson St, Ann Arbor, MI 48104, USA. E-mail: [email protected] Received 26 May 2016; revised version accepted 00 0000

ABSTRACT Background and Objectives: Non-communicable diseases may reflect an evolutionary mismatch between our human ancestry and modern environments. To explore the mismatch hypothesis for Acne vulgaris, we studied the prevalence and severity of acne in Dogon adolescents in Mali, West Africa. Methodology: We graded the prevalence and severity of acne in 1182 Dogon adolescents aged 11–18 years from nine villages using facial photos taken as part of a prospective cohort study. Eighty-nine (89%) of the individuals in the cohort migrated to the city during adolescence, enabling us to assess the effect of urban migration. Using multivariable logistic regression, we estimated the effect of predictor variables on the presence of acne. Results: The prevalence of acne in the cohort was 28%, with 90% of cases being mild or very mild. Thus, the prevalence and severity of acne was much lower than for adolescents in high-income countries. Controlling for age, puberty, and body mass index (BMI), the odds of boys developing acne was 85% lower in the city than in the villages (P = 0.002). Conclusion and Implications: Acne is similar to the ‘diseases of civilization’ in being promoted by the pro-inflammatory properties of modern diets. The low prevalence and severity of acne in the Dogon supports the mismatch hypothesis and suggests that acne should join the list of diseases of modern lifestyles. However, we also observed an unexpected decrease in acne in urban boys. Future research is needed for a deeper mechanistic understanding of the interplay between diet, inflammation, immune function and other environmental exposures that differ between urban and rural environments.

INTRODUCTION significant perception of disfigurement [3,7], In high-income countries, Acne vulgaris is one of the decreased self-esteem [7], depression [8] and lower most common skin diseases [1–5] and one of the prospects for employment [9]. Acne vulgaris is a most frequent diagnoses for patients who visit a multifactorial disease [10,11] that is influenced by dermatologist [6]. Although not life threatening, four agreed upon factors: (1) hyperkeratinization the lesions and scarring of Acne vulgaris can cause and obstruction of sebaceous follicles caused by

ß The Author(s) 2016. Published by Oxford University Press on behalf of the Foundation for Evolution, Medicine, and Public Health. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. 326 | Campbell and Strassmann Evolution, Medicine, and Public Health

abnormal desquamation of the follicular epithelium; Here we report the prevalence and severity of acne (2) androgen-stimulated increase in sebum produc- in Dogon adolescents in Mali, West Africa. We tion, (3) colonization of hair follicles in the skin by hypothesized a low prevalence and severity of acne Propionibacterium acnes, and (4) inflammation in the Dogon on account of their traditional life-style, [4,5,12–14]. The pathogenesis of acne is similar including a diet based on millet farming with across different ethnic populations [6] and acne limited consumption of novel, pro-inflammatory, has been documented within all Fitzpatrick skin fast foods. Inflammation plays an important role in types (The Fitzpatrick scale is a numerical classifica- the development of acne [27] and in the aetiology of tion of skin color from I to VI with Type I being the other chronic diseases [28] associated with modern fairest (albino, very fair with freckles) to Type VI lifestyles. Examples include cardiovascular disease being the darkest (deeply pigmented, 0black0 skin)) [29,30] type 2 diabetes mellitus [29], Alzheimer’s dis- [6,15]. Here we consider Acne vulgaris within the ease [31], and various cancers [32]. Modern dietary framework of evolutionary medicine. habits promote inflammation, especially consump- Human bodies are imperfectly adapted for a var- tion of foods that have a high glycemic load or a high iety of reasons including the ‘mismatch’ between ratio of omega-6 to omega-3 polyunsaturated fatty our genes and novel features of the environment acids (PUFAs) [28]. The evidence that pro-inflamma- [16]. However, there was no singular Environment tory foods simultaneously exacerbate acne [4] and of Evolutionary Adaptedness or specific time-period other chronic diseases [28], raises the possibility that of our evolutionary past to which humans are Acne vulgaris (especially inflammatory acne) should adapted [17–19]. Instead different aspects of human be added to the list of ‘diseases of civilization’. biology evolved in different time periods [19,20]. The We also hypothesized that acne in the Dogon is aggravated by urban migration—a globally import- early emphasis on Stone Age genes was unwar- ant source of environmental novelty in the 21st cen- ranted [21] as many genes are far older (e.g. Hox tury. The world’s population is >50% urban and will genes) and others are more recent than the be 66% urban by 2050 [33]. The Dogon are migrating Pleistocene (e.g. genes for lactase persistence) to Bamako, which is the fastest growing city in Africa [20,22]. Nor was the advent of agriculture the only and the sixth fastest growing city in the world [34]. disease-inducing transition in the human past [19]. When rural adolescents from farming families move Environmental and lifestyle changes relevant to dis- to the city, they are introduced to a radically different ease risk occurred both long before and long after environment compared to the one they experienced the end of the Stone Age [19]. during development. In the city, one might expect a Although it is helpful to put the Stone Age in per- more acnegenic diet due to increased consumption spective, discarding the mismatch hypothesis is un- of pro-inflammatory foods such as red meat [35] milk justifiable. When the environment changes, traits [36], food with a high glycemic index [12, 37–40], or a that were adaptive in the previous environment be- high ratio of omega-6 to omega-3 fatty acids [42–46]. come maladaptive in the new environment until nat- One might also expect increased psychosocial stress ural selection has ‘caught up’. Quite simply, natural due to loss of kin support or increased competition selection takes time to act [23–25]. A striking ex- from wealthy urbanites. Further, in the city people ample was proposed by Janzen and Martin (1982) tend to be more sedentary and do not get the stress- [26], who studied the peculiarly large and hard fruits reducing benefits of aerobic farm work [47]. of many trees in the tropical forests of Central Previous studies reported a low prevalence of America. These fruits appear to be anachronistically acne in traditional populations. For example, the adapted for dispersal by gomphotheres (a relative of prevalence of acne was 0% in the Kitavan Islanders the mastodon with powerful molar teeth) and other of Papua New Guinea and the Ache´ foragers of large mammals that went extinct about 10 000 years Paraguay [4], 6.9% in adolescents of a roadside ago. Ten millennia has not been long enough for the Ethiopian community [48], and 4.2% in adolescents evolution of fruits that will be eaten and dispersed by in a cross-sectional, hospital-based study in Mali smaller mammals, so these overly-large and overly- [49]. Furthermore, several studies reported that the protected fruits rot under their parent trees [24,26]. Inuit [50] and Chinese [51] only developed acne after Similarly, our study of acne is predicated on the transitioning to a modern lifestyle. However, these understanding that after the environment changes, studies [4, 48] had very small sample sizes of ado- it will take time for natural selection to catch up. lescents: from 47 to 300 individuals (Table 1). Our h lmse fmdr society? modern of blemishes The

Table 1. Studies of adolescent acne prevalence by country

Background of Community Studied Country Age (years) N Total Moderate to Referenceb Prevalence (%) Severe (% of Total Prevalence) Subsistence horticulture, ﬁsherman Papua New Guinea 15–25 300 00 Cordain et al. 2002 Roadside town of 4000, subsistence farming Ethiopia 10–16 47 2.7 NRa Figueroa et al. 1996 Millet farming, rural villages, Mali 11–18 1182 28.4 9.2 This study & city of 2 million 10 rural villages of Kakamega county Kenya 10–19 133 38.3 N/A Kiprono et al. 2015 Students in town of Arequipa, Peru 12–18 2214 41.7 19.0 Freye et al.1998 population 800,000 Secondary school in Nottingham, United Kingdom 14–16 317 50.0 27.4 Smithard et al. 2001 population 310,000 Eskisehir province (developed) and Turkey 13–18 2230 60.0 21.4 Aksu et al. 2011 surroundings (rural) Randomly selected secondary schools New Zealand 12–18 9398 67.3 20.9 Purvis et al. 2004 Secondary schools in state of Victoria Australia 13–18 797 83.5 21.0 Kilkenny et al. 1998 Secondary schools and colleges Singapore 13–19 1045 88.0 48.6 Tan et al. 2007 Schools in Kaduna, population 1.5 million Nigeria 11–19 418 90.7 7.3 Yahya 2009 Students in Tehran, population 8 million Iran 12–20 1,002 93.2 15.0 Ghodsi et al. 2009 Students in Antwerp, population 500,000 Belgium 14–18 594 95.0 38.0 Nijsten et al. 2005 aNR= no results. b Grading scale information: (1) Cordain 2002: 3 grades: mild, moderate, severe; (2) Figuerora et al. 1996: presence of acne ‘Yes’ or ‘No’; (3) This study: 5 point scale (see Table 1); (4) Kiprono et 327 al. | Strassmann and Campbell 2015: not available; (5) Freye et al. 1998: 4 point scale: minimal, mild, moderate severe*; (6) Smithard et al. 2001: 3 grades: very mild, mild, moderate to severe; (7) Aksu et al. 2011: 4 grades: minimal, mild, moderate, severe*; (8) Purvis et al. 2004: 3 grades: None, Some, Severe; (9) Kilkenny et al. 1998: 4 point scale: minimal, mild, moderate, severe; (10) Tan et al. 2007: 3 grades: mild, moderate, severe; (11) Nijsten et al. 2005: 3 grades: little, moderate, severe; (12) Yahya 2009: GAGS 30 point scale*; (13) Ghodsi et al. 2009: 3 grades: mild, moderate, severe; *refer to paper for speciﬁc criteria. 328 | Campbell and Strassmann Evolution, Medicine, and Public Health

study is the largest (N = 1182) study of adolescents METHODOLOGY in a traditional, rural population and it has the ad- The subjects are participants in a prospective cohort vantage of being population rather than hospital or study initiated by BIS in 1998. In May 1998 all chil- school-based [52]. Whereas previous studies dren age 0–5 years in 9 villages were enrolled in the controlled for few covariates—at most age or age study and, during the following two years, all new- and body mass index (BMI) [53,54], we explored borns were added. Thus, the data are for the total multivariable models in which we tested the simul- population of children (N = 1698) in the specified taneous effects of age, BMI, puberty, wealth and age range in the 9 villages. By 2012, 245 children urban versus rural living on presence or absence of had died (14.5%), 1427 were alive (84.2%), and 23 acne. Furthermore, our study design is unique in (1.4%) were lost to follow-up. Among the living chil- that we compared urban and rural members of the dren, 127 had moved to the capital of Mali, Bamako, same cohort from nine source villages, thereby and a similar number had moved to other cities, reducing other causes of heterogeneity between in- including Abidjan, in Cote d’Ivoire. In the present dividuals. The members of our study population had study, we examined the facial identity photographs the same surname, reducing genetic variability, and taken between November 2011 and April 2012 for since they came from villages that were within 5 km of each other, cultural variability was minimal. 1263 adolescents (88% of the survivors), including 89 adolescents who had moved to the capital city of Bamako. A total of 1182 facial photos, or 94% of the Ethnographic background total, were of sufficient quality to be graded for acne. The Dogon villages we studied are on the These 1182 adolescents ranged in age from 11.5 to Bandiagara Escarpment in central Mali and retain 18.8 years (mean age 14.8 ± 2.0) and consisted of some of the salient features of the human evolution- 554 females and 628 males (46.9% and 53.1%, re- ary past including the absence of contraception, spectively). A total of 1093 of our subjects lived in the high fertility and mortality, polygyny and life in a rural villages whereas 42 boys and 48 girls were living tight-knit web of kin [52,55]. The Dogon are farmers in the city at the time of evaluation. The images were whose main subsistence crop is pearl millet enlarged and viewed via Windows Photo Viewer by (Pennisetum glaucum), an ancient West African grain CEC. [56] that is usually eaten twice per day. Other local The grading system consisted of a 5-point scale cereals include sorghum (Sorghum bicolor), fonio developed by Dreńo et al. (2011) [59](Table 2). (Digitaria exilis) and rice (Oryza glaberrima)[57]. We chose this scale because of its wide range of Mangos, oranges, bananas, papayas, guavas, grades (very mild to very severe), which improved tomatoes and wild fruits are part of the diet; whereas the precision of classification. The high amount of vegetables, such as eggplant, play a minor role with melanin in the skin of these adolescents made an the exception of onions–the primary cash crop. Dairy assessment of the level of inflammatory lesions dif- products are rarely consumed although canned or ficult. Thus the major determinant for acne powdered milk is added to Nescafe´, a luxury item. grade was the facial surface area with acne lesions, Locally grown peanuts are commonly eaten as though we did note the presence of inflammatory snacks and beans are another source of protein. In lesions. wealthier families, adult men consume morsels of Acne lesions and acne severity are generally fresh meat (beef, mutton, goat, pork or chicken). divided into two categories: non-inflammatory and Refined sugar is added to porridge and beverages inflammatory. Non-inflammatory lesions are small like wild fruit drinks, Nescafe´, Lipton and Arabic 1–2 mm comedones, commonly known as white tea. Palm oil often substitutes for the traditional oil or black heads, whereas inflammatory lesions are from the shea plant (Vitellaria paradoxa). Aside from 5–6 mm papules, pustules, or nodules [2,3]. palm oil and the small amount of meat, dairy, rice Likewise, non-inflammatory acne, also known as and refined sugar in the Dogon diet, most foods are comedonal acne, consists of smaller, less severe le- non-inflammatory and have a low glycemic load. sions [2,3], and inflammatory acne consists of larger, Thus, Dogon food stuffs contrast with the pro-in- more inflamed lesions. Despite the differentiation flammatory, processed foods in modern diets (e.g. between non-inflammatory and inflammatory le- white bread, soda, processed snacks) [58] (See SI sions, some dermatologists attest that all acne le- Discussion). sions are inflamed to varying degrees [27]. The blemishes of modern society? Campbell and Strassmann | 329

Table 2. Deﬁnition of acne grade and severity. Source: Dre´no et al. (2011). Scale for determination of acne severity in Dogon adolescents

Grade Severity Description 0 Clear, No lesions Residual pigmentation and erythema may be seen. 1 Very Mild A few scattered open or closed comedones and very few papules. 2 Mild Easily recognizable: less than half the face involved. A few open or closed comedones and/or a few papules and pustules 3 Moderate More than half the face involved. Many open or closed comedones and/or many papules and pustules. 4 Severe Entire face involved, covered with many papules and pustules and/or many open or closed comedones and rare nodules. 5 Very Severe Highly inﬂammatory acne covering the face with presence of nodules.

In our study, evidence of at least one non-inflam- Institutional (HSBS) Review Board (HUM00013631). matory lesion (comedone) qualified as ‘very mild’ Permission to score the existing photos for acne in a acne (grade 1). Other grading systems do not con- way that protected anonymity was determined to be sider the presence of one non-inflammatory lesion exempt from IRB regulation (HUM00076491) on 5/10/ to be acne [2,60]. However, since comedones are 2013 by the HSBS IRB at the University of Michigan. often precursors to more severe acne and are evidence for some level of acne within the population, we considered it preferable to use a grading system Statistical methods that was conservative. Thus, our methodology We ran Spearman rank correlations [65] between our classifies more individuals as having Acne vulgaris binary acne variable and each of our candidate pre- than might be the case for other grading schemas dictor variables. We then created multivariable logis- [2,60–62]. Due to the difficulty of accurately detect- tic regression models, in which the dependent ing and diagnosing scars and hyperpigmentation variable was the presence or absence of acne and caused by acne, we did not consider these two fac- the predictors were age, BMI, pubertal development, tors in our classification. We created a binary acne wealth and urban vs. rural living (Table 3). We variable, defined as follows: 0 = absent, 1 = present analyzed the data separately for each sex as several (any grade) because of the high prevalence of no variables were pertinent only to girls (breast stage, acne. menarche and buttocks circumference) or boys A female Dogon research assistant visually in- (e.g., testosterone). All analyses were carried out spected the girls’ breasts without palpation and using IBM SPSS (Statistical Package for the Social scored them according to Tanner breast stage [63]. Sciences) software (versions 18–22) [66]. We con- A girl was considered to be pubertal if she had sidered the results to be statistically significant if reached at least stage 2 as this breast stage is P < 0.05. defined as the onset of thelarche (breast budding) and is the marker for puberty [64]. For girls, we gathered data on menarche (status quo) and RESULTS buttocks circumference (cm), and for boys, we assayed saliva samples to measure testosterone Acne severity levels (pg/ml). Each subject’s height (cm), weight (kg), BMI (kg/m2), percent body fat, abdominal skin- The prevalence of acne by grade for the 1182 Dogon fold thickness (mm), subscapular skinfold thickness adolescents was as follows: no acne: 71.8% (mm), waist circumference (cm) and wealth rank (z- (N = 849), very mild: 15.3% (N = 182), mild: 10.4% score) was measured. Location of residence (urban (N = 123), moderate: 2.5% (N = 29), severe: 0.2% or rural) was also recorded. (N = 2) and very severe: 0% (N =0)(Fig. 1). Cases Approval for the fieldwork was obtained from the of moderate acne did not occur until age 13 (3.1% of Malian government and the University of Michigan all 13 year olds) and increased to 11.9% by age 18 Health Sciences and Behavioural Sciences years (Fig. 2A, Supplementary Material, Table S1). 330 | Campbell and Strassmann Evolution, Medicine, and Public Health

Table 3. Multivariable logistic regression results with acne (0 = not present, 1 = present) as the dependent variable; and age, BMI, puberty, wealth and urban versus rural residence as predictor variables

Model 1 Girls Boys

Exp (B) 95% C.I. P Exp (B) 95% C.I. P Age (years) 1.253 1.057–1.485 0.009 1.426 1.205–1.688 0.000 BMI (kg/m2) 1.129 1.017–1.253 0.023 1.323 1.119–1.565 0.001 Puberty (Girls: 0 = no, 1 = yes; 2.187 1.144–4.184 0.018 1.013 1.004–1.023 0.006 Boys: Testosterone (pg/ml)) Wealth (z-score) 0.933 0.736–1.183 0.566 0.992 0.762–1.291 0.951 Urban living (no, yes) 0.578 0.208–1.608 0.293 0.148 0.044–0.505 0.002

Figure 1. Number of Dogon adolescents (age 11–18) with each acne grade (percentage of population above bar)

Acne prevalence

Using our binary acne variable (0 = absent, 1 = present), the presence of acne increased progressively for the entire cohort from 6.6% at age 11 to 56.7% at Figure 2. Percentage of population with acne (any grade). (A) by severity and age (years); (B) by sex and age (years) age 18 (Supplementary Material, Table S1). The overall prevalence across all ages was 26.4% in males and 30.5% in females (Supplementary had acne, 17.6% had at least one inflammatory le- Material, Table S1). The distribution by age was sion and 82.4% had no inflammatory lesions. similar for boys and girls, although the increase with age occurred earlier in girls (Fig. 2B; Supplementary Spearman rank correlation coefficients: Material, Table S1). bivariate results Acne (any severity) was present in 23.6% of girls who had not reached menarche and in 51.1% who In both sexes, acne presence (having any acne) was had reached menarche. The prevalence of acne significantly positively correlated with age, BMI, increased with Tanner breast stage (Fig. 3A) in girls height, weight, subscapular skinfold, abdominal skin- and with testosterone level in boys (Fig. 3B). Acne fold and waist circumference. In girls, acne presence prevalence increased with BMI until BMI was additionally correlated with buttock circumfer- reached >25 kg/m2, after which the sample size ence, tanner breast stage, and menarche. In boys, was small (Fig. 3C). Among the individuals who acne presence was correlated with testosterone level The blemishes of modern society? Campbell and Strassmann | 331

increased with age. Specifically, for a 1 year increase in age, the odds of having acne increased 25% in girls (P = 0.009) and 43% (P = 0.000) in boys. For a one unit (kg/m2) increase in BMI, the odds of acne increased 13% in girls (P = 0.023) and 32% (P= 0.001) in boys. For females, having a Tanner Breast stage of two or higher (indicative of having reached puberty) increased the odds of having acne by a factor of 2.2 (P = 0.018), and in males, a 10 pg/ml increase in testosterone increased the odds of having acne 13% (P=0.060). Wealth was not significantly related to acne presence in either sex. Urban living was not significantly associated with the acne presence in girls; however, in boys, moving to the capital city, Bamako, decreased the odds of having any acne by 85% (P = 0.002) (Table 3).

DISCUSSION

Variation in acne prevalence and severity

Acne occurs in the Dogon adolescent population (age 11.5–18.8) at a fairly low prevalence (28.3%) and severity, with the majority of cases (90.8%) being ‘very mild’ or ‘mild’ (Fig. 1, Table 4). In addition, the prevalence would be nearly cut in half (to 13.1%) if the lowest acne grade (very mild) were not Figure 3. Percentage of population with acne (any grade). considered acne—consistent with the grading (A) by Tanner breast stage (1 - 5); (B) by testosterone level method for three previous studies [61,67,68]. The (pg/ml); (C) by BMI (kg/m2) low severity of acne suggests that the majority of Dogon adolescents had non-inflammatory acne. (pg/ml). Percent body fat was significantly negatively Although detecting inflammatory lesions was diffi- correlated with acne presence in boys and signifi- cult, only 17.6% of Dogon adolescents with acne of cantly positively correlated with acne in girls. Wealth any grade had 1 inflammatory lesion. In addition, (z-score) was not significantly correlated with acne we observed no other severe form of acne, such as prevalence for boys or girls (girls: Spearman’s Acne conglobata [69,70] r =0.006, P =0.849; boys: Spearman’s r = 0.031, The low prevalence of acne in the Dogon contrasts P =0.452) (Supplementary Material, Table S2A with the higher prevalence reported for high income and B). In the bivariate results, living in an urban countries or urban areas of middle income countries setting was weakly, positively correlated with acne where lifestyles are presumably more modern. For presence for both boys and girls although not quite example, the prevalence of acne in adolescents living significantly so (girls: Spearman’s r = 0.075, in Singapore [62], Australia [61], United Kingdom P = 0.079; boys: Spearman’s r = 0.075, P=0.060) [67], Belgium [71] and New Zealand [68] ranged from (Supplementary Material, Table S2A and B). 50 to 95%, with 21 to 47% of individuals with acne having a grade of moderate to severe (Table 1). Acne prevalence in urban communities of middle-income Logistic regressions: multivariable results countries such as Peru [72], Iran [1], Turkey [73] and The multivariable logistic regression results are Nigeria [74], ranged from 42% to 93%. Among the shown in Table 3. Descriptive statistics for the vari- individuals who had acne in these countries, 7.3% to ables in the model, broken down by gender (female/ 21.4% had the moderate to severe grades (Table 1). male) and residence (urban/rural), are shown in Thus, acne grade was higher (more severe) in the Table 4. In both sexes, the odds of having acne high than in the middle income countries and both 332 | Campbell and Strassmann Evolution, Medicine, and Public Health

Table 4. Descriptive statistics (mean, standard deviation and sample size) by gender and rural versus urban residence; acne prevalence; and acne grade in Dogon adolescents.

Female Male

Rural Urban Rural Urban

N 506 48 586 42

Mean SD Mean SD Mean SD Mean SD Age (years) 14.6 1.85 16.4 1.99 14.6 1.97 16.6 1.96 BMI (kg/m2) 17.5 2.72 21.3 3.21 16.9 1.82 18.8 2.24 Puberty (Girls: 0 = no, 1 = yes; 0.6 0.49 0.9 0.32 33.2 25.20 86.9 55.19 Boys: Testosterone (pg/ml)) Wealth (z-score) 0.5 0.90 0.9 0.90 0.5 0.92 0.6 0.60 ACNE % N % N % N % N Prevalence (any grade) 29.4 149 41.7 20 25.6 150 38.0 16 Mean SD Mean SD Mean SD Mean SD Grade (0–5) 0.5 0.81 0.7 0.99 0.4 0.74 0.7 0.93

acne grade and prevalence were conspicuously low Ghana. Furthermore, the Dogon had less acne than in the Dogon, as we originally hypothesized. did Africans in urban environments, except Accra. Despite the low prevalence of acne in Dogon adolescents overall, we note that the prevalence did Causes of variation in acne prevalence & reach 56.7% at age 18 years, a percentage similar severity (see SI discussion) to urban Peruvians (city: Arequipa) of the same age (59.4%) [72], but lower than the 93.3% preva- Lifestyle factors may explain the low prevalence of lence observed at ages 16–18 years in Australia [61] acne in the Dogon, including a low glycemic load and the 97.1% prevalence observed at ages 17–19 diet [4,12, 37–40] and limited milk consumption years in urban Nigeria [74]. Although 56.7% of [36]. The Dogon diet consists of food with glycemic Dogon adolescents had acne at age 18, the majority indices as low as 1.7 (peanuts), 6.8 (mangoes) and of these cases (88.1%) were very mild or mild. The 16.8 (millet), which contrasts sharply with the gly- only studies that found the prevalence of acne to be cemic indices found in the traditional western diet lower than that of the Dogon were conducted in trad- (i.e. 34.7 for white bread and 64.9 for refined sugar) itional, non-urban populations [4,48,75](Table 1). [4,58,79]. Multiple randomized controlled trials have Acne prevalence in African communities varies found that a high glycemic load diet can increase the greatly: from 0.1% in rural Tanzania (unspecified risk for acne [37–39]. Nonetheless, other aspects of ages) [76], to approximately 4% in rural Cameroon the Dogon lifestyle such as sweetening hot drinks (ages 10–19) [75], and 38% in rural Kenya (ages 10– with refined sugar, consuming palm oil and psycho- 19) [77]. In urban populations, acne prevalence social stress may contribute to the acne that we did reached 65% in Durban, the second most populous observe (see SI Discussion). urban area in South Africa [78], for individuals 13–24 years of age; and 97% for individuals aged 17–19 in Predictors of acne in the Dogon urban Nigeria [74]. In a study that compared both rural and urban communities in Ghana, 0.2% of Three inter-related variables, age (years), BMI rural school children had acne (N = 527) compared (kg/m2) and puberty (Testosterone pg/ml for boys to 12.9% (N = 489, P<0.001) in the capital, Accra and Tanner Breast stage 2 for girls) were signifi- [53]. The Dogon have a lower acne prevalence than cantly and positively associated with the presence the rural Kenyans but a higher prevalence than the of acne in Dogon adolescents, even when included rural populations of Tanzania, Cameroon and in the same statistical model (Table 3). The positive The blemishes of modern society? Campbell and Strassmann | 333 correlation with age is consistent with previous It is unclear why there was no difference in the studies [72,80–82] and is probably mediated by an- odds of having acne by rural versus urban residence drogenic hormones released during puberty [83]. for girls. The lifestyles of boys and girls in the city The finding that puberty for both girls and boys differ—with girls likely to work as maids and boys was positively correlated with acne replicates previ- more likely to be enrolled in high school (lyceés). ous studies [80–83]. Lastly, the positive and highly However, when we added a variable for attending significant correlation between acne presence and school (no, yes) to the multivariable model it was BMI was expected since BMI increases during pu- not statistically significant for either sex (P = 0.68 for berty [82]. Other studies have reported acne to be boys and P = 0.57 for girls) and it did not affect the associated with a high BMI (overweight and obesity) odds ratio or p-value for urban residence or the other for girls [54] however, none have had sufficient data covariates in the model. Specifically, the reduction of to evaluate individual predictors of acne when acne prevalence in urban boys was 86% (P = 0.004) controlling for others in a multivariable model—as after adjusting for school in the model in Table 3.In we have done (Table 3). ongoing research, we are investigating the effect of The odds ratios of our findings with respect to urban migration and gender on stress as measured puberty in boys and girls are not directly comparable by blood pressure, cortisol levels and the global to each other. We used a threshold measure for pu- perceived stress scale [86]. It will be helpful to know berty in girls (breast stage two or higher) and a con- how urban migration affects stress levels in both tinuous variable for testosterone level (pg/ml) in sexes. boys. A one unit increase in testosterone (pg/ml) Our finding of lower odds of acne in urban boys is is much less than the difference between having consistent with two studies in Turkey and Portugal. breast stage I versus II or higher in girls. In these studies, the prevalence of acne in semi-rural or low-income students was higher than in students living in an urban environment [73,87]. The authors Wealth of the Turkish study speculated that this difference was caused by increased access to specialized med- Interestingly, we found that wealth was not signifi- ical doctors, such as dermatologists, in the city [87]. cantly correlated with acne development. Our ex- This explanation is unlikely to apply to the Dogon pectation had been that a richer diet in wealthier who, regardless of urban or rural residence, rarely families would lead to an increase in acne. It is pos- visit specialists. sible that wealthier adolescents consumed more We note that 42 boys (6.7%) lived in the city and sugar, milk and rice, which are potential risk factors 586 (93.3%) lived in the rural villages. Likewise, only for acne [4,12, 37–40] but were also better nourished. 48 girls (8.7%) lived in the city compared to 506 Thus, there may have been opposing forces: better (91.3%) in the villages (Table 4). Although our sam- nutrition and immune function in the wealthier ple size for urban adolescents was limited, we had a families counterbalanced by eating more foods with strong effect size and a highly significant P-value for a high glycemic index like sugar. It is also possible the urban-rural comparison in boys (Odds that there wasn’t enough variation in diet to observe ratio = 0.85, P = 0.002). a correlation between acne and wealth. The reduced odds of acne in urban Dogon boys did not show up in the Spearman rank correlation test. In this bivariate test, living in an urban setting Urban vs. rural was weakly positively correlated with acne for both Another significant and unexpected finding is that boys and girls, although not quite significantly so urban living was associated with an 85% decrease (girls: Spearman’s r = 0.075, P=0.079; boys: (P = 0.002) in the odds of having any acne for boys. Spearman’s r=.075, P =.06) (Supplementary The underlying mechanisms are unclear. It is pos- Material, Table 2A and B). A likely reason for the sible that urban boys had access to a more nutritious positive correlation in bivariate analyses is that the diet that included a greater variety of vegetables that subjects who moved to the city were older, on aver- provide vitamins that support the immune system. age, than the subjects who remained in the rural area In ongoing research, we are examining the effect of (Table 4). Moreover, age at puberty was earlier for gender and urban migration on C-reactive protein, Dogon adolescents who moved to the city (Table 4). which is a marker for inflammation [84,85]. The decrease in acne risk for urban boys showed up 334 | Campbell and Strassmann Evolution, Medicine, and Public Health

only in our multivariable models in which age, BMI, environments. Using a carefully controlled study de- puberty and wealth were controlled. Thus bivariate sign in which we compared members of the same studies of acne prevalence in urban versus rural cohort who migrated to the city against those who settings may be misleading. Importantly, the sub- stayed in their rural villages, boys in the city actually jects we measured in the city were from the same had lower odds of having acne. In contrast with pre- rural villages as the subjects we measured in the vious studies, we controlled for BMI, puberty and rural areas, increasing the validity of our rural-urban wealth. These results suggest that a more fine-tuned comparison. examination of the interplay between health, immune function, diet, inflammation, stress and environmental exposures is needed to shed light on acne Study limitations risk (SI Discussion). Dichotomies like traditional In our study, the acne grader (CEC) was not a versus modern, or urban versus rural, mask the certified dermatologist, however advice on acne underlying variation in risk factors that contribute grading was provided by two practicing dermatolo- to the aetiology of acne. gists (Gerard Plewig MD of the University of Munich, Germany and Lisa Maier MD, clinician at the acknowledgements University of Michigan Health System). The photos of each adolescent were not taken specifically for We thank Gerald Plewig MD and Lisa Maier MD for help with acne grading purposes and hence lighting condi- scoring acne; Helen Dixon PhD for helpful comments; Annie tions, though similar, were not identical. Due to Xiao for help creating the figures; and Dan Kruger PhD for the high amount of melanin in the subjects’ skin, help with the title. We are grateful to the Dogon adolescents accurate detection of comedones and inflammation for participating in this study and to the Malian government was challenging. Finally, comparison of acne preva- for authorizing it. lence and severity in the Dogon to that of other populations was made more difficult by the multiplicity of funding grading scales employed (Table 1)–to date, there is no universally agreed upon grading scale for acne. The fieldwork was supported by National Science Foundation Furthermore, puberty in the Dogon occurs rela- grant BCS 1029056 to BIS. tively late, with girls reaching menarche about five years later than in the United States. As the age range supplementary data of our subjects was 11.5 to 18.8 years, we cannot report on acne prevalence after age 19 years. Other Supplementary data are available at EMPH online. studies of adolescent acne in African populations, Conflict of interest: None declared. had age ranges similar to those in our study: 10–16 years in Ethiopia [48], 10–19 years in Kenya [77] and references 11–19 years in Nigeria [74]. We note that the prevalence of acne may be slightly underestimated when 1. Ghodsi S, Orawa H, Zouboulis C. Prevalence, severity, members of a late-maturing population (like the severity risk factors of acne in high school pupils: a com- Dogon and other African populations) are compared munity-based study. J Invest Dermatol 2009;129:2136–41. by age to members of early maturing populations in 2. Plewig G, Kligman AM. 2000 Acne and Rosacea. Berlin & Australia or Europe. Heidelberg: Springer,. 3. Layton AM, Thiboutot D, Vincenzo B. 2004 Acne. Oxford, UK: Health Press. CONCLUSION AND IMPLICATIONS 4. Cordain L, Lindeberg S, Hurtado M et al. Acne vulgaris: a In comparison to studies of adolescents living in disease of Western Civilization. Arch Dermatol 2002;38: 1584–90. high income countries or urban areas, the overall 5. Degitz K, Placzek M, Borelli C et al. Pathophysiology of prevalence of Acne vulgaris in the Dogon was low acne. J Dtsch Dermatol Ges 2007;4:316–23. doi: 10.1111/ and the majority of cases were mild or very mild j. 1610-0387. (90%). These results are consistent with the hypoth- 6. Alexis A. Acne in Patients With Skin of Color. J Drugs esis that acne is a disease that is aggravated by mod- Dermatol 2011; 10(suppl):13–6. ern lifestyles. However, this hypothesis predicts that 7. Magin P, Adams J, Heading G et al. The causes of acne: a acne should be higher in urban than in rural qualitative study of patient perceptions of acne causation The blemishes of modern society? Campbell and Strassmann | 335

and their implications for acne care. Dermatol Nurs 28. Galland L. Diet and inflammation. Nutr Clin Pract 2006;18:344–9. 2010;25:634–40. 8. Hull P, D’arcy C. Acne, depression, and suicide. Clin 29. Rubio-Ruiz M, Peredo-Esca´rcega A, Cano-Martinez A et al. Dermatol 2005;23:665–74. An Evolutionary Perspective of Nutrition and 9. Cunliffe WJ. Acne and unemployment. Br J Dermatol Inflammation Mechanisms of Cardiovascular Disease. 1986;115:386. Int J Evol Biol 2015; doi.org/10.1155/2015/179791 10. Cordain 2. Implications in the role of diet in acne. Semin 30. Cordain L, Eaton S, Sebastian A et al. Origins and evolu- Cutan Med Surg 2005;24:84–91. tion of the Western diet: health imlications for the 21st 11. Melnik B. Milk Consumption: aggravating factor of acne century. Am J Clin Nutr 2005;81:341–54. and promoter of chronic diseases in Western societies. J 31. Daniel G, Arnold S. Some evolutionary perspectives on Dtsch Dermatol Ges 2009;4:364–70. doi: 10.1111/j.1610- Alzheimer’s disease pathogenesis and pathology. 0387.2009.07019. Alzheimer’s and Dementia 2012;8:343–51. 12. Bowe W, Whitney P, Joshi S et al. Diet and acne. Am Acad 32. Hochberg M, Thomas F, Assenat E. Preventive evolution- Dermatol 2010;63:124–41. ary medicine of cancers. Evol Appl 2012; 134–43. 13. Leyden J. New understandings of the pathogenesis of 33. United Nations 2016. UN World’s population inceasingly acne. J Am Acad Dermatol 1995;32(suppl:15S–25S. urban with more than half living in urban areas http:// 14. Gollnick H, Cunliffe W, Berson D et al. Management of www.un.org/en/development/desa/news/population/ Acne: A Report From a Global Alliance to Improve world-urbanization-prospects-2014.html (19 August Outcomes in Acne. J Am Acad Dermatol 2003; 49 2016, date last accessed). (suppl:S1–38). 34. Index of Statistics 2016. http://www.citymayors. 15. Davis E, Callender V. A review of acne in ethnic skin. J Clin com/statistics/urban_growth1.html(19 August 2016, date Aesthetic Dermatol 2010;3:24–38. last accessed). 16. Nesse R, Williams G. 1999 16–22. Research designs that 35. Samraj A, Pearce O, Laubli H et al. A red meat-derived address evolutionary questions about medical disorders. glycan promotes inflammation and cancer progression. In: Stearns, S (ed. Evolution in Health and Disease. New Proc Natl Acad Sci U S A 2015;112:542–7. York: Oxford University Press Inc, 36. Adebamowo CA, Spiegelman D, Danby FW et al. High 17. Foley R. The adaptive legacy of human evolution: A search school dietary dairy intake and teenage acne. J Am Acad for the environment of evolutionary adaptedness. Evol Dermatol 2005;52:207–14. Anthropol 1995;4:194–205. 37. Smith RN, Braue A, Varigos GA et al. The effect of a low 18. Irons W. Adaptively Relevant Environments Versus the glycemic load diet on acne vulgaris and the fatty acid com- Environment of Evolutionary Adaptedness. Evol position of skin, surface triglyceride. J Dermatol Sci 2005; Anthropol 1998;6:194–204. 50:41–52. 19. Strassmann BI, Dunbar R. 1999 91–100. Human evolution 38. Smith RN, Mann NJ, Braue A et al. A low glycemic-load & disease: putting the stone age in perspective. In: diet improves symptoms in acne vulgaris patients: a Stearns, S (ed. Evolution in Health and Disease. New randomizedcontrolledtrial.Am J Clin Nutr 2007; 86:107–15. York: Oxford University Press Inc, 39. Smith RN, Mann N, Makelainen H et al. A pilot study to 20. Caroll S. 2005 Endless forms most beautiful. New York: W.W. determine the short-term effects of a low glycemic load Norton & Company,. diet on hormonal markers of acne: a nonrandomized, par- 21. Eaton SB, Konner M, Shostak M. Stone agers in the fast allel controlled feeding trail. Mol Nutr Food Res lane: chronic degenerative diseases in evolutionary per- 2008;52:718–26. spective. Am J Med 1988;84:739–49. 40. Ingram JR, Grindlay JC, Williams HC. Management of acne 22. Tishkoff S, Reed F, Ranciaro A et al. Convergent adaptation vulgaris: an evidence-based update. Clin Exp Dermatol of human lacaste persistence in Africa and Europe. Nat 2009; 35:351–4. Genet 2007;31–40. 42. Jung JY, Kwon HH, Hong JS et al. Effect of dietary supple- 23. Darwin C. 1859 On the Origin of Speices. London: John mentation with omega-3 fatty acid and gamma-linolenic Murray. acid on ace vulgaris: a randomised, double-blind, 24. Ridley M. 1993 Evolution. Cambridge: Blackwell Scientific controlled trial. Acta Derm Venereol 2014;94:521–5. Publications. 43. Simopoulos AP. Omega-3 fatty acids in inflammation and 25. Stearns S. 1999 3–15. Introducing evolutionary thinking. autoimmune diseases. J Am Coll Nutr 2002;21:495–505. Stearns, S (ed. Evolution in Health and Disease. New York: 44. Zouboulis C, Seltmann H, Hiroi N et al. Corticotropin- Oxford University Press Inc. releasing hormone: An autocrinehormone that promotes 26. Janzen DH, Martin PS. Neotropical anachronisms: the lipogenesis in human sebocytes. Proc Natl Acad Sci U S A fruits the gomphotheres ate. Science 1982; 215:19–27. 2002;99:7148–53. 27. Harvey A, Huynh T. Inflammation and acne: putting the 45. Kraus S. Stress, acne and skin surface free fatty acids. pieces together. J Drugs Dermatol 2014;13:459–63. Psychosom Med 1970;32:503–8. 336 | Campbell and Strassmann Evolution, Medicine, and Public Health

46. Ganceviciene R, Bohm M, Fimmel S et al. The role of 64. Davison K, Susman E, Birch L. Percent body fat at age 5 neuropeptides in the multifactorial pathogenesis of acne predicts earlier pubertal development among girls at age vulgaris. Dermatoendocrinol 2009;1:170–6. 9. Pediatrics 2003; 111:815–21. 47. De Geus EJC. 2010 73–8. Aerobic Exercise and Stress 65. Motulsky H. 1995 Intuitive Biostatistics. New York: Oxford Reduction. In: Encyclopedia of Stress, doi: 10.1016/ University Press. B978-012373947-6.00014-3. 66. IMB Corp 2012. IBM SPSS Statistics for Windows, version 48. Figueroa J, Fuller C, Abraha A et al. The prevalence of skin 18-22. Armonk, NY: IBM Corp. disease among school children in rural ethiopia – a pre- 67. Smithard A, Glazebrook C, Williams H. Acne prevalence, liminary assessment of dermatologic needs. Pediatr knowledge about acne and psychological morbidity Dermatol 1996;13:378–81. in mid-adolescence: a community-based study. Br J 49. Mahe´ A, Cisse´ I, Faye O et al. Skin diseases in Bamako Dermatol 2001; 145:274–97. (Mali). Int J Dermatol 1998;37:673–6. 68. Purvis D,Robinson E, Watson P. Acneprevalence in second- 50. Schaefer O. When the Eskimo comes to town. Nutr Today aryschoolstudentsandtheirperceiveddifficultyinaccessing 1971;6:8–16. acne treatment. NZMedJ 2004; 117:1–8. 51. Steiner P. Necropsies on Okinawans: anatomic and patho- 69. Petros H. What’s your assessment? Dermatol Nurs 2007; logic observations. Arch Pathol 1946;42:359–80. 19:368–70. 52. Strassmann BI. Cooperation and competition in a cliff- 70. Shirakawa M, Uramoto K, Harada F. Treatment of acne dwelling people. Proc Natl Acad Sci U S A 2011;108: conglobate with infliximab. J Am Acad Dermatol 2006; 10894–901. 55:344–6. 53. Hogewoning AA, Koeleij I, Amoah AS et al. Prevalence and 71. Nijsten T, Rombouts S, Lambert J. Acne is prevalent but risk factors of inflammatory acne vulgaris in rural and use of its treatment is infrequent among adolescents from urban Ghanaian schoolchildren. Br J Dermatol 2009;161: the general population. J Eur Acad Dermatol Venereol 470–92. doi: 10.1111/j.1365-2133.2009.09259.x ... 2007;21:163–8. 54. Halvorsen J, Vleugels R, Bjertness E et al. A population- 72. Freye E, Rebaza R, Sami D et al. The prevalence of facial based study of acne and body mass index in adolescents. acne in peruvian adolescents and its relation their ethni- Arch Dermatol 2012;148:131–2. city. J Adolesc Health 1998;22:480–4. 55. Strassmann BI. 2000 49–67. Polygyny, family structure, 73. Aksu AE, Metintas S, Saracoglu ZN et al. Acne: prevalence and child mortality: A prospective study among the and relationship with dietary habits in Eskisehir, Turkey. Dogon of Mali. In: Cronk L, Chagnon N, Irons W (eds) J Eur Acad Dermatol Venereol 2012;12:1503–9. doi: Adaptation and Human Behavior: An Anthropological 10.1111/j.1468-3083.2011.04329.x Perspective. New York: Aldine de Gruyter, 74. Yahya H. Acne vulgaris in Nigerian adolescents – preva- 56. National Research Council 1996. Lost Crops of Africa Volume 1: lence, severity, beliefs, perceptions, and practices. Int J Grains. Washtingon D.C.,National Academy Press. Dermatol 2009;48:498–505. 57. Strassmann B, Warner J. Predictors of fundability and con- 75. Bissek A, Tabah E, Koutou E et al. The spectrum of skin ception waits among the Dogon of Mali. Am J Phys diseases in rural setting Cameroon (sub-Saharan Africa). Anthropol 1998;105:167–84. BMC Dermatol 2012;12:7. 58. Foster-Powell K, Holt S, Brand-Miller J. International table 76. Gibbs S. Skin disease and socioeconomic conditions in of glycemic index and glycemic load values: 2002. Am J Clin rural Africa: Tanzania. Int J Dermatol 1996; 35:633–9. Nutr 2002; 76:5–56. 77. Kipronos S, Wamburu G. Acne vulgaris in general popu- 59. Dreńo B, Foli F, Pawin H et al. Development and evalu- lation of rural western Kenya: cross-sectional community ation of a Global Acne Severity Scale (GEA Scale) suitable survey. Intl Soc Dermatol 2016;55:e211–26. for France and Europe. J Eur Acad Dermatol Venereol 78. Dlova NC, Mankahla A, Madala N et al.Thespectrumofskin 2011;25:43–8. diseases in a black population in Durban, KwaZulu-Natal, 60. Cunliffe WJ, Gould DJ. Prevalence of facial acne vulgaris South Africa. Intl J Dermatol 2015;54: 279–85. in late adolescence and in adults. Br Med J 1979;1: 79. Nutrition Data 2013: Self: Estimated Glycemic Load. http:// 1109–10. nutritiondata.self.com/help/estimated-glycemic-load (12 61. Kilkenny M, Merlin K, Plunkett A et al. The prevalence of June 2016, date last accessed) common skin conditions in Australian school students: 3 80. Lucky AW, Biro FM, Huster GA et al. Acne vulgaris in early Acne vulgaris. Br J Dermatol 1998; 139:840–5. adolescent boys: correlations with pubertal maturation 62. Tan HH, Tan AW, Barkham T et al. Community-based and age. Arch Dermatol 1991; 127:210–6. study of acne vulgaris in adolescents in Singapore. Br J 81. Lucky A, Biro F, Huster GA et al. Acne Vulgaris in Dermatol 2007; 157:547–51. Premenarchal Girls. Arch Dermatol 1994;30:308–14. 63. Tanner JM. 1962 Growth at adolescence: with a general con- 82. Burton JL, Cunliffe WJ, Stafford I et al. The prevalence of acne sideration of the effects of hereditary and environmental fac- vulgaris in adolescence. Br J Dermatol 1971; 85:119–26. tors upon growth and maturation from birth to maturity. 83. Arora M, Yadav A, Saini V. Role of hormones in acne Oxford: Oxford University Press. vulgaris. J Clin Biochem 2011;44:1035–40. The blemishes of modern society? Campbell and Strassmann | 337

84. Libby P, Ridker P. Inﬂammation and Atherosclerosis: Role 86. Cohen S, Kamarck T, Mermelstein R. A global meas- of C-Reactive Protein in Risk Assessment. Am J Med 2004; ure of perceived stress. JHealthSocBehav 1983;24: 116:9S–16S. 385–96. 85. Shrivastava A, Singh H, Raizada A et al. C-reactive protein, 87. Amado JM, Breu AM, Loureiro L et al. The prevalence of inﬂammation and coronary heart disease. Egyptian Heart J acne in the north of Portugal. J Eur Acad Dermatol Venereol 2014; 67:89–97. 2006;20:1287–95.